一种mongodb集群数据布局优化方法研究-中国云计算.pdfVIP

Computer Engineering and Applications 计算机工程与应用 2017 ，53（17） 77 一种MongoDB 集群数据布局优化方法研究 用 1，2 1，2 1，2 1，2 冯东煜 ，朱立谷 ，肖子达 ，刘 迪 应 FENG Dongyu1，2 , ZHU Ligu1，2 , XIAO Zida1，2 , LIU Di1，2 与 1. 中国传媒大学 计算机学院，北京 100024 程 2. 安防大数据处理与应用北京市重点实验室，北京 100094 g r 1. College of Computer, Communication University of China, Beijing 100024, China 工 o 2. Beijng Key Laboratory of Big Data in Security Protection Industry, Beijing 100094, China . 机 j a FENG Dongyu, ZHU Ligu, XIAO Zida, et al. Approach for optimizing data placement on MongoDB cluster. 算 e Computer Engineering and Applications, 2017, 53 （17）：77-84. c 计 w. Abstract: Traditional relational database exposed many issues difficult to overcome in dealing with large-scale data appli- cation, NoSQL is widely used under the background of big data due to its unique characteristics. It takes data application w of posting and delivering in express industry as research background, to study the optimization approach of data place- w ment in MongoDB cluster. First it introduces posting and delivering off- line data analytic system based on MongoDB. Then it researches the shard key strategy of MongoDB on the basis of fields ’characteristics, proposes a striped placement approach for continuous uniform data based MongoDB shard tag. Finally it tests the performance of data placement approach, the results show that MongoDB cluster with this approach has better data di